An activity tracking device, such as a step-counting device includes an interface configured to receive one or more acceleration signals and signal processing circuitry. The signal processing circuitry generates an indication of condition of an accelerometer, such as a body position of the accelerometer, based on one or more accelerometer signals, generates an event signal, such as an event flag, based on one or more accelerometer signals and the indication of the condition of the accelerometer, and generates an activity signal, such as step flag based on the event signal, the indication of the condition of the accelerometer and one or more acceleration signals. The signal processing circuitry may generate a noise signal based on one or more acceleration signals and generate the activity signal based on the noise signal.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method, comprising: generating an indication of a condition related to an accelerometer based on one or more accelerometer signals and one or more statistical classifiers trained using training data, wherein the indication of the condition related to the accelerometer is an indication of a body position of the accelerometer; generating an event signal based on the one or more accelerometer signals and the indication of the body position of the accelerometer; and generating an activity signal based on the event signal, the indication of the body position of the accelerometer and the one or more accelerometer signals.
This invention relates to accelerometer-based activity monitoring, specifically detecting body position and activity events. The method processes accelerometer signals using statistical classifiers trained on historical data to determine the accelerometer's body position, such as whether it is attached to a limb, torso, or other body part. The system then analyzes the accelerometer signals in conjunction with the detected body position to identify specific events, such as movement patterns or impacts. These events are further processed with the accelerometer signals and body position data to generate an activity signal, which may represent a user's physical activity, posture, or other motion-related metrics. The statistical classifiers are trained to recognize patterns in accelerometer data that correlate with different body positions and activities, enabling accurate event detection and activity classification. This approach improves the reliability of activity monitoring by accounting for the accelerometer's placement and orientation on the body. The system is useful for applications like fitness tracking, medical monitoring, or industrial safety, where understanding body position and movement is critical.
2. The method of claim 1 , comprising: generating a noise signal based on the one or more accelerometer signals; and generating the activity signal based on the noise signal.
A method for processing accelerometer signals to generate an activity signal involves detecting and analyzing motion data from one or more accelerometers. The method addresses the challenge of accurately identifying and classifying physical activities or movements using accelerometer data, which can be affected by noise and variability in sensor readings. The process begins by capturing accelerometer signals, which represent motion data from one or more sensors. These signals are then processed to generate a noise signal, which isolates or characterizes the noise components present in the raw accelerometer data. The activity signal is subsequently derived from this noise signal, enabling more accurate detection and classification of activities such as walking, running, or other movements. The method may also include preprocessing steps to filter or enhance the accelerometer signals before noise signal generation. By leveraging the noise signal, the approach improves the reliability and precision of activity recognition systems, particularly in environments where sensor noise or interference is a concern. The technique is applicable in wearable devices, fitness trackers, and other motion-sensing applications where accurate activity monitoring is essential.
3. The method of claim 1 wherein the activity signal is a step signal and the method comprises generating a step count signal based on the step signal.
This invention relates to activity monitoring systems, specifically methods for processing activity signals to generate step count data. The technology addresses the challenge of accurately detecting and quantifying human movement, particularly walking or running steps, from raw activity signals. The method involves analyzing an activity signal, which is a step signal representing detected physical activity, and generating a step count signal based on this input. The step count signal provides a numerical representation of the number of steps taken, enabling applications such as fitness tracking, health monitoring, or activity analysis. The method may include preprocessing the activity signal to filter noise or enhance relevant features before step detection. The step count signal can be used in various devices, including wearable sensors, smartphones, or medical monitoring systems, to provide users with real-time or historical step data. The invention improves upon existing step detection techniques by offering a more reliable and accurate way to process step signals, ensuring precise step counting for better activity tracking.
4. The method of claim 3 , comprising: generating a noise signal based on the one or more accelerometer signals; and generating the step count signal based on the noise signal and the step signal.
This invention relates to a method for improving step counting accuracy in wearable devices using accelerometer data. The problem addressed is the inherent noise and inaccuracies in step detection algorithms when relying solely on accelerometer signals, which can lead to false positives or missed steps. The method enhances step counting by incorporating noise signal analysis alongside traditional step detection techniques. The method involves processing one or more accelerometer signals to generate a noise signal, which represents the non-step-related motion or environmental interference captured by the accelerometer. This noise signal is then used in conjunction with a step signal, which is derived from conventional step detection algorithms, to produce a refined step count signal. By analyzing the noise signal, the method can distinguish between true steps and false detections caused by vibrations, sudden movements, or other non-walking activities. The integration of noise signal analysis improves the reliability of step counting, particularly in dynamic or noisy environments. The method may also include preprocessing the accelerometer signals to filter out irrelevant frequencies or artifacts before generating the noise signal. The step signal is typically generated using threshold-based or pattern-matching techniques applied to the accelerometer data. The final step count signal is derived by combining the step signal with the noise signal, either through weighted averaging, statistical analysis, or machine learning models trained to differentiate between true steps and noise. This approach reduces errors in step counting, making it more accurate for fitness tracking, medical monitoring, or other applications requiring precise motion analysis.
5. The method of claim 1 wherein the event signal comprises an event flag, and the method comprises generating the activity signal based on one or more acceleration data blocks associated with the event flag.
This invention relates to systems for processing event signals, particularly in applications involving motion or activity detection. The problem addressed is the need to accurately generate an activity signal from event data, such as acceleration measurements, to detect and analyze physical events or movements. The method involves using an event signal that includes an event flag, which marks the occurrence of a relevant event. The system processes one or more acceleration data blocks associated with this event flag to generate an activity signal. The acceleration data blocks contain motion information, such as changes in velocity or position, which are analyzed to determine the nature and characteristics of the event. The activity signal is then derived from this analysis, providing a representation of the detected activity. The method ensures that only relevant acceleration data blocks, those linked to the event flag, are used in generating the activity signal. This improves accuracy by filtering out irrelevant or noisy data, ensuring that the activity signal reliably reflects the actual event. The approach is useful in applications like wearable devices, industrial monitoring, or any system requiring precise event detection and analysis based on motion data.
6. The method of claim 1 , comprising: controlling an application program based on the activity signal.
A system and method for controlling an application program based on user activity involves detecting and analyzing user activity signals to dynamically adjust or modify the behavior of an application. The method includes monitoring user interactions, such as gestures, movements, or physiological signals, to generate an activity signal representing the user's state or intent. This activity signal is then processed to determine appropriate control actions for the application, such as adjusting settings, triggering functions, or modifying content display. The system may use sensors, such as cameras, motion trackers, or biometric devices, to capture user activity data. The activity signal may be derived from patterns in the user's behavior, such as frequency, intensity, or context of movements. The application program is then controlled in real-time based on the interpreted activity signal, enabling adaptive and responsive interactions. This approach enhances user experience by making applications more intuitive and context-aware, particularly in scenarios where traditional input methods are impractical or inefficient. The method may be applied in various domains, including gaming, accessibility, virtual reality, and smart environments, where dynamic control based on user activity improves functionality and usability.
7. The method of claim 1 wherein the generating the event signal based on the one or more accelerometer signals and the indication of the body position of the accelerometer comprises selecting event trigger criteria based on the indication of the body position.
The invention relates to a method for generating event signals from accelerometer data, particularly in systems where the position of the accelerometer on a body affects signal interpretation. The problem addressed is accurately detecting events (e.g., impacts, movements) from accelerometer signals when the sensor's placement on the body influences how the data should be processed. For example, an accelerometer on a wrist may require different event detection criteria than one on a waist. The method involves analyzing accelerometer signals and the body position of the sensor to generate an event signal. The key improvement is dynamically selecting event trigger criteria based on the body position. For instance, if the accelerometer is on a limb, the system may use higher sensitivity to detect subtle movements, whereas if it is on the torso, it may apply stricter thresholds to filter out background noise. This ensures accurate event detection regardless of sensor placement. The method may also involve preprocessing the accelerometer signals (e.g., filtering, normalization) before applying the position-based criteria. The result is a more reliable event detection system adaptable to different body positions.
8. The method of claim 1 wherein the generating the activity signal based on the event signal, the indication of the body position of the accelerometer and the one or more accelerometer signals comprises selecting a trained statistical classifier from a plurality of trained statistical classifiers based on the indication of the body position.
This invention relates to activity monitoring systems that use accelerometer data to detect and classify human activities. The problem addressed is accurately identifying activities from accelerometer signals, which can vary significantly depending on the position of the sensor on the body. For example, walking detected on the wrist differs from walking detected on the ankle, requiring different classification models. The method involves generating an activity signal from an event signal and one or more accelerometer signals, where the accelerometer is attached to a user's body. The system first determines the body position of the accelerometer, such as whether it is on the wrist, ankle, or another location. Based on this position, the system selects a trained statistical classifier from a plurality of classifiers. Each classifier is specifically trained for the detected body position, improving accuracy by using a model optimized for the sensor's location. The selected classifier processes the accelerometer signals and event signal to generate an activity signal, such as identifying walking, running, or other movements. This approach ensures that the classification is tailored to the sensor's placement, reducing errors caused by positional variations. The system may also include preprocessing steps to condition the accelerometer signals before classification.
9. A device, comprising: an interface configured to receive acceleration signals: and acceleration signal processing circuitry coupled to the interface, and which, in operation: generates an indication of a condition related to an accelerometer based on one or more acceleration signals and one or more statistical classifiers trained using training data, wherein the indication of the condition related to the accelerometer is an indication of a body position of the accelerometer; generates an event signal based on the one or more acceleration signals and the indication of the condition related to the accelerometer; and generates an activity signal based on the event signal, the indication of the condition related to the accelerometer and the one or more acceleration signals.
The invention relates to a device for analyzing acceleration signals from an accelerometer to determine body position and activity. The device includes an interface to receive acceleration signals and processing circuitry that evaluates these signals using statistical classifiers trained on prior data. The classifiers identify the accelerometer's body position, such as whether it is in a stationary, moving, or specific orientation state. The device then generates an event signal based on the acceleration data and the detected body position. Additionally, it produces an activity signal derived from the event signal, the body position, and the raw acceleration data. This activity signal may represent a user's movement patterns or other dynamic conditions. The statistical classifiers are trained to recognize patterns in the acceleration data that correlate with specific body positions, enabling accurate real-time or post-processing analysis. The device is useful in applications like fitness tracking, medical monitoring, or industrial motion analysis, where understanding body position and activity is critical. The system improves upon traditional accelerometer analysis by incorporating machine learning to enhance accuracy and contextual awareness.
10. The device of claim 9 wherein the acceleration signal processing circuitry, in operation, generates a noise signal based on the one or more acceleration signals; and generates the activity signal based on the noise signal.
The invention relates to a device for processing acceleration signals to generate an activity signal, addressing the challenge of accurately detecting and analyzing physical activity or motion using sensor data. The device includes acceleration signal processing circuitry that receives one or more acceleration signals from one or more accelerometers. The circuitry processes these signals to generate a noise signal, which represents variations or disturbances in the acceleration data. The activity signal is then derived from this noise signal, enabling the device to identify and quantify physical activity or motion patterns. The noise signal may be generated by filtering, amplifying, or otherwise transforming the acceleration signals to isolate relevant activity-related components. The activity signal can be used for applications such as fitness tracking, motion analysis, or health monitoring, where distinguishing meaningful motion from background noise is critical. The device may also include additional components, such as a sensor interface for receiving the acceleration signals and a communication interface for transmitting the activity signal to an external system. The overall system ensures reliable activity detection by leveraging noise-based signal processing techniques.
11. The device of claim 9 wherein the activity signal is a step signal and the signal processing circuitry, in operation, generates a step count signal based on the step signal.
The invention relates to a wearable or portable device designed to monitor and analyze physical activity, specifically focusing on step detection and counting. The device includes signal processing circuitry that receives an activity signal, which in this case is a step signal generated by sensors such as accelerometers or gyroscopes. The signal processing circuitry processes this step signal to generate a step count signal, which quantifies the number of steps taken by the user. This functionality is part of a broader system that may also include additional sensors and processing modules to track other types of physical activity or physiological metrics. The step counting feature is particularly useful for fitness tracking, rehabilitation monitoring, or general health and wellness applications, providing users with accurate and real-time feedback on their movement patterns. The device may be integrated into smartwatches, fitness bands, or other wearable technologies, ensuring portability and continuous monitoring. The step detection algorithm within the signal processing circuitry is optimized to filter out noise and distinguish between actual steps and other movements, improving accuracy in various environments and activities.
12. The device of claim 11 wherein the acceleration signal processing circuitry, in operation, generates a noise signal based on the one or more acceleration signals; and generates the step count signal based on the noise signal and the step signal.
The invention relates to a device for processing acceleration signals to improve step counting accuracy. The device addresses the problem of inaccurate step detection in wearable or motion-tracking devices, particularly in noisy environments where false positives or missed steps occur. The device includes acceleration signal processing circuitry that receives one or more acceleration signals from an accelerometer. The circuitry generates a noise signal derived from the acceleration signals, which represents unwanted motion or interference. The circuitry then combines this noise signal with a step signal, which is a preliminary step detection output, to produce a refined step count signal. This refinement process reduces errors caused by noise, ensuring more reliable step counting. The device may also include additional components, such as a sensor interface for receiving acceleration data and a processor for executing step detection algorithms. The noise signal generation and step signal refinement improve the accuracy of step counting in dynamic or noisy conditions, making the device suitable for fitness trackers, medical monitoring systems, or other motion-sensing applications.
13. The device of claim 9 , comprising: a plurality of data buffers, wherein the event signal comprises an event flag, and the acceleration signal processing circuitry, in operation, generates the activity signal based on one or more acceleration data blocks associated with the event flag and stored in the plurality of data buffers.
The invention relates to a device for processing acceleration signals, particularly in systems where event-based data collection is used to monitor physical activity or motion. The problem addressed is efficiently processing acceleration data triggered by specific events, such as motion detection, to generate meaningful activity signals while minimizing computational overhead and power consumption. The device includes acceleration signal processing circuitry that receives an event signal containing an event flag, which indicates the occurrence of a relevant event. The circuitry processes acceleration data blocks associated with this event flag, which are stored in a plurality of data buffers. The buffers temporarily hold acceleration data until the processing circuitry generates an activity signal based on the buffered data. This approach allows for selective processing of acceleration data only when an event occurs, reducing unnecessary computations and improving energy efficiency. The data buffers ensure that acceleration data is retained until the processing circuitry can analyze it, preventing data loss during high-activity periods. The activity signal, derived from the buffered acceleration data, provides a consolidated output representing the detected motion or activity. This method is particularly useful in wearable devices, IoT sensors, or other applications where power efficiency and real-time processing are critical. The invention optimizes resource usage by focusing processing efforts on event-triggered data rather than continuously analyzing all incoming acceleration signals.
14. The device of claim 9 , comprising: an application processor, which, in operation, controls an application program based on the activity signal.
A system for monitoring and controlling application programs based on user activity includes a sensor module that detects user activity and generates an activity signal. The system also includes an application processor that controls an application program in response to the activity signal. The application processor may adjust the application program's behavior, such as pausing, resuming, or modifying its operation, based on the detected activity. The sensor module may include one or more sensors, such as motion sensors, proximity sensors, or biometric sensors, to detect various types of user activity. The system may be integrated into a computing device, such as a smartphone, tablet, or wearable device, to optimize power consumption and user experience by dynamically adapting application behavior to the user's current activity state. The application processor may also prioritize certain applications or functions based on the activity signal, ensuring that critical operations continue while non-essential functions are suspended or reduced. This system enhances efficiency by reducing unnecessary processing and improving responsiveness to user needs.
15. The device of claim 9 wherein the acceleration signal processing circuitry, in operation, generates the event signal based on the one or more acceleration signals and the indication of the condition related to the accelerometer by selecting event trigger criteria based on the indication of the condition.
The invention relates to an accelerometer-based device designed to detect and process acceleration signals while accounting for operational conditions that may affect accelerometer performance. The device includes acceleration signal processing circuitry that generates an event signal based on one or more acceleration signals and an indication of a condition related to the accelerometer. The circuitry dynamically selects event trigger criteria based on the indicated condition, ensuring accurate event detection even when the accelerometer's performance is influenced by factors such as temperature, mechanical stress, or calibration drift. This adaptive approach improves reliability in applications where environmental or operational conditions vary, such as in industrial monitoring, automotive systems, or wearable devices. The device may also include additional components like a sensor interface for receiving acceleration signals and a communication interface for transmitting the event signal to external systems. The adaptive selection of trigger criteria allows the device to maintain consistent performance across different operating scenarios, reducing false positives or missed detections. This solution addresses the challenge of maintaining accurate event detection in accelerometer systems subjected to varying conditions, enhancing overall system robustness.
16. The device of claim 9 wherein the acceleration signal processing circuitry, in operation, generates the activity signal based on the event signal, the indication of the condition related to the accelerometer and the one or more acceleration signals by selecting a statistical classifier from a plurality of statistical classifiers based on the indication of the condition.
This invention relates to a device for processing acceleration signals to generate an activity signal, addressing challenges in accurately detecting and classifying activities using accelerometer data. The device includes an accelerometer that generates one or more acceleration signals representing motion, and acceleration signal processing circuitry that processes these signals to produce an activity signal indicating a detected activity. The circuitry also receives an event signal and an indication of a condition related to the accelerometer, such as sensor noise, calibration state, or environmental factors, which can affect signal quality. The processing circuitry selects a statistical classifier from a plurality of classifiers based on the condition indication to improve activity detection accuracy. The selected classifier analyzes the acceleration signals and event signal to generate the activity signal, ensuring robust performance under varying conditions. The device may also include a memory storing the classifiers and a processor executing instructions to perform the classification. This approach enhances reliability in applications like fitness tracking, medical monitoring, or industrial motion analysis by dynamically adapting to sensor conditions.
17. The device of claim 9 wherein the signal processing circuitry comprises: one or more filters configured to filter received acceleration signals.
The invention relates to signal processing in devices that measure acceleration, addressing the challenge of accurately extracting meaningful data from raw acceleration signals. The device includes signal processing circuitry designed to enhance the quality of acceleration measurements by filtering out noise and irrelevant frequency components. The circuitry comprises one or more filters that process received acceleration signals to isolate relevant motion data. These filters may include low-pass, high-pass, or band-pass filters, depending on the application, to remove unwanted signal artifacts while preserving the desired motion information. The filtered signals can then be used for applications such as motion tracking, vibration analysis, or inertial navigation. The filtering step ensures that the processed acceleration data is more reliable and suitable for further analysis or control systems. The device may be part of a larger system, such as a wearable sensor, industrial monitoring equipment, or a navigation system, where accurate acceleration measurements are critical. The filters are configured to adapt to different environmental conditions or operational requirements, improving the overall performance of the device.
18. The device of claim 17 wherein the one or more filters comprise at least one of: a low-pass filter; and a high-pass filter.
This invention relates to a device for processing signals, particularly in applications where filtering is required to isolate or remove specific frequency components. The device includes one or more filters designed to modify the frequency content of an input signal. The filters may include a low-pass filter, which allows low-frequency components to pass while attenuating higher frequencies, or a high-pass filter, which permits high-frequency components to pass while blocking lower frequencies. These filters can be used individually or in combination to achieve desired signal processing outcomes, such as noise reduction, signal conditioning, or frequency separation. The device is particularly useful in electronic systems where precise control over signal frequency content is necessary, such as in audio processing, telecommunications, or sensor signal conditioning. The inclusion of both low-pass and high-pass filtering options provides flexibility in adapting the device to different applications, ensuring that the output signal meets specific performance requirements.
19. The device of claim 9 , comprising: an accelerometer coupled to the interface and configured to generate the one or more acceleration signals.
A wearable device monitors physical activity by detecting and analyzing movement patterns. The device includes a sensor interface that collects data from one or more motion sensors, such as accelerometers, to generate acceleration signals. These signals are processed to determine movement characteristics, such as speed, direction, and intensity, which are then used to assess physical activity levels. The device may also include additional sensors, such as gyroscopes or magnetometers, to enhance motion tracking accuracy. The collected data is transmitted to a processing unit, which analyzes the signals to identify specific activities, such as walking, running, or cycling. The device may further include a display or notification system to provide real-time feedback to the user. The accelerometer is directly coupled to the sensor interface, ensuring precise and reliable signal generation for accurate activity monitoring. This system enables users to track their physical activity, set fitness goals, and receive personalized recommendations based on their movement patterns. The device is particularly useful for fitness enthusiasts, athletes, and individuals seeking to improve their overall health through regular physical activity.
20. A method, comprising: generating an indication of a condition related to an accelerometer based on one or more accelerometer signals, wherein the condition is a body position of the accelerometer; selecting event trigger criteria based on the indication of the condition related to the accelerometer; generating an event signal based on the one or more accelerometer signals, the indication of the condition related to the accelerometer and the selected event trigger criteria; and generating an activity signal based on the event signal, the indication of the condition related to the accelerometer and the one or more accelerometer signals.
The invention relates to accelerometer-based activity monitoring systems, specifically addressing the challenge of accurately detecting and classifying human movement or body position using accelerometer data. Traditional systems often struggle with distinguishing between different activities or body positions due to variations in sensor orientation and movement patterns. This invention provides a method to improve activity detection by dynamically adjusting event trigger criteria based on the accelerometer's body position. The method involves generating an indication of the accelerometer's body position by analyzing one or more accelerometer signals. This position data is then used to select appropriate event trigger criteria, which define the conditions under which an event (such as a step, fall, or other movement) is detected. An event signal is generated by processing the accelerometer signals in combination with the body position indication and the selected event trigger criteria. Finally, an activity signal is produced by further analyzing the event signal, the body position indication, and the raw accelerometer signals to classify the detected activity. By dynamically adjusting event detection based on body position, the system improves accuracy in identifying and classifying activities, reducing false positives and enhancing reliability in applications such as fitness tracking, medical monitoring, or safety systems.
21. The method of claim 20 , comprising: generating a noise signal based on the one or more accelerometer signals; and generating the activity signal based on the noise signal.
The invention relates to systems and methods for processing accelerometer signals to generate an activity signal, particularly in applications where noise reduction and accurate activity detection are important. The method involves analyzing one or more accelerometer signals to identify and extract relevant motion data while suppressing noise. A noise signal is generated from the accelerometer signals, which is then used to refine the activity signal, ensuring that only meaningful motion data is retained. This approach improves the accuracy of activity detection by distinguishing between true motion events and background noise. The method may be applied in wearable devices, fitness trackers, or medical monitoring systems where reliable activity tracking is essential. By dynamically adjusting the noise signal, the system adapts to varying environmental conditions, enhancing performance in real-world scenarios. The technique ensures that the activity signal accurately reflects the user's movements while minimizing false positives from irrelevant vibrations or disturbances. This solution addresses challenges in motion sensing, particularly in noisy environments, by providing a robust method for isolating and processing motion data.
22. The method of claim 20 wherein the activity signal is a step signal and the method comprises generating a step count signal based on the step signal.
The invention relates to activity monitoring systems, specifically methods for processing activity signals to generate step count data. The problem addressed is accurately detecting and quantifying human movement, particularly steps, from raw activity signals. Traditional methods often struggle with noise, irregular movements, or varying step patterns, leading to inaccurate step counts. The method processes an activity signal, which is a step signal representing detected human movement. The step signal is analyzed to generate a step count signal, which quantifies the number of steps taken. This involves filtering and interpreting the raw activity data to distinguish true steps from other movements or noise. The step count signal provides a reliable metric for tracking physical activity, useful in fitness devices, medical monitoring, or research applications. The method ensures accurate step detection by refining the step signal before counting, improving reliability over existing approaches that may miscount or miss steps. The system may integrate with wearable or embedded sensors to capture the activity signal, enabling real-time or post-processing step analysis. The invention enhances activity tracking by providing a more precise step count, addressing limitations in prior art systems that rely on less sophisticated signal processing techniques.
23. The method of claim 20 wherein the event signal comprises an event flag, and the method comprises generating the activity signal based on one or more acceleration data blocks associated with the event flag.
This invention relates to systems for processing event-based signals, particularly in applications involving motion or activity detection. The problem addressed is the need to accurately generate an activity signal from raw sensor data, such as acceleration data, in a way that is synchronized with specific events of interest. The method involves capturing acceleration data in discrete blocks and associating these blocks with an event flag, which marks a significant event or condition. The system then processes the acceleration data blocks linked to the event flag to generate an activity signal. This signal represents the detected motion or activity, allowing for further analysis or triggering of subsequent actions. The approach ensures that the activity signal is derived from relevant data segments, improving accuracy and reducing noise from unrelated motion. The method may also include filtering or analyzing the acceleration data blocks to enhance signal quality before generating the activity signal. By focusing on data blocks tied to the event flag, the system efficiently isolates meaningful motion events from background noise, making it suitable for applications like wearable devices, industrial monitoring, or environmental sensing. The technique ensures that the activity signal is both timely and contextually relevant to the detected event.
24. The method of claim 20 , comprising using a statistical classifier to generate the indication of a body position.
A method for determining body position using statistical classification techniques. The method involves analyzing input data, such as sensor measurements or images, to detect and classify a person's body position. A statistical classifier, such as a machine learning model, processes the input data to generate an indication of the body position, which may include standing, sitting, lying down, or other postures. The classifier is trained on labeled datasets containing examples of different body positions to improve accuracy. The method may also involve preprocessing the input data to enhance feature extraction and reduce noise. The output of the classifier is used to provide feedback, trigger actions, or monitor health conditions based on the detected body position. This approach is useful in applications like healthcare monitoring, ergonomic assessments, and assistive technologies where accurate body position detection is critical. The statistical classifier may be implemented using algorithms such as decision trees, support vector machines, or neural networks, depending on the specific requirements of the application. The method ensures reliable and real-time body position detection, improving user experience and system performance in various domains.
25. A device, comprising: an interface configured to receive acceleration signals: and acceleration signal processing circuitry coupled to the interface, and which, in operation: generates an indication of a condition related to an accelerometer based on one or more acceleration signals, wherein the condition is a body position of the accelerometer; selects event trigger criteria based on the indication of the condition related to the accelerometer; generates an event signal based on the one or more acceleration signals and the indication of the condition related to the accelerometer; and generates an activity signal based on the event signal, the indication of the condition related to the accelerometer and the one or more acceleration signals.
The device is designed for monitoring and analyzing acceleration data to determine body position and activity events. The system includes an interface to receive acceleration signals from an accelerometer and processing circuitry that evaluates these signals to detect the accelerometer's body position. Based on this position, the device selects specific event trigger criteria to identify relevant motion events. The processing circuitry then generates an event signal by analyzing the acceleration data in relation to the detected body position. Finally, the device produces an activity signal that combines the event signal, the body position indication, and the raw acceleration data to provide a comprehensive assessment of the accelerometer's movement and orientation. This approach enables accurate detection of activities and events by dynamically adjusting criteria based on the accelerometer's position, improving reliability in applications such as wearable devices, fitness tracking, or industrial monitoring. The system enhances the precision of motion analysis by integrating positional context with acceleration data.
26. The device of claim 25 wherein the acceleration signal processing circuitry, in operation, generates a noise signal based on the one or more acceleration signals; and generates the activity signal based on the noise signal.
The invention relates to a device for processing acceleration signals to detect and analyze physical activity. The device includes acceleration signal processing circuitry that receives one or more acceleration signals from one or more accelerometers. The circuitry generates a noise signal derived from the acceleration signals, which represents variations or disturbances in the measured acceleration data. The circuitry then processes this noise signal to produce an activity signal, which indicates the presence, type, or characteristics of physical activity being performed by the user. This approach allows for the detection of subtle movements or vibrations that may not be apparent in raw acceleration data, improving the accuracy of activity monitoring. The device may be part of a wearable or portable system designed for fitness tracking, medical monitoring, or industrial applications where motion analysis is required. The noise signal generation step involves filtering, amplification, or other signal conditioning techniques to isolate relevant activity-related components from the acceleration data. The resulting activity signal can be used to classify activities, trigger alerts, or provide feedback to the user. The invention enhances the reliability of motion-based activity detection by leveraging noise signal analysis.
27. The device of claim 25 , comprising: an accelerometer coupled to the interface and configured to generate the one or more acceleration signals.
A device for monitoring physical activity or motion includes an interface that receives one or more acceleration signals from an accelerometer. The accelerometer is coupled to the interface and generates the acceleration signals, which represent motion or movement data. The device may be used in applications such as fitness tracking, industrial monitoring, or medical diagnostics, where accurate motion detection is required. The accelerometer provides real-time or continuous data, allowing the device to analyze movement patterns, detect anomalies, or trigger specific actions based on the detected acceleration. The interface processes the signals to extract relevant information, such as speed, direction, or impact forces, enabling precise motion tracking and analysis. The accelerometer may be a microelectromechanical system (MEMS) sensor or another type of motion-sensing component, depending on the application. The device ensures reliable motion detection by integrating the accelerometer directly with the interface, minimizing signal loss or distortion. This configuration enhances accuracy and responsiveness, making it suitable for high-performance applications where precise motion data is critical.
28. The device of claim 25 , wherein the acceleration signal processing circuitry, in operation, uses one or more statistical classifiers to generate the indication of a body position.
The invention relates to a device for monitoring body position using acceleration signals. The device addresses the challenge of accurately determining a person's body position, such as standing, sitting, or lying down, based on motion data. The device includes acceleration signal processing circuitry that analyzes acceleration data to detect changes in body orientation and movement patterns. This circuitry employs statistical classifiers, such as machine learning models or pattern recognition algorithms, to interpret the acceleration signals and classify the body position. The classifiers are trained to recognize distinct acceleration patterns associated with different body positions, improving accuracy in real-world conditions. The device may also include additional sensors, such as gyroscopes or magnetometers, to enhance position detection. The statistical classifiers may use features like signal amplitude, frequency components, or temporal patterns to distinguish between positions. The output is an indication of the current body position, which can be used for applications like health monitoring, fitness tracking, or assistive technologies. The invention improves upon prior systems by leveraging advanced statistical techniques to handle variability in movement and environmental factors.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
June 6, 2019
March 22, 2022
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.